Quarterly data from 1962 to 1983 for beef, lamb, mutton, pig meat and poultry were used to test for constancy in the structure of meat demand in Australia. The cumulative sum, cumulative sum of squares and Farley-Hinich tests were applied to a range of models to ensure that any rejection of stability was not due to an inappropriate functional form or omitted dynamics. Little evidence was found of a marked swing away from consumption of any meat, with the exception of mutton. The results suggest that changes in prices and in total consumer expenditure are far more important than changes in tastes as determinants of meat consumption.